Context: Locating products or services on-line that meet users want is more and more troublesome thanks to the massive pool of decisions to think about before inbound at the required one. A user might pay a substantial quantity of your time exploring varied on-line resources to find things that work his needs. Moreover, users might not forever categorical their preferences in an exceedingly manner that simply matches them to things that might meet them. Objective: Searching out things on-line has been done principally through information queries that come back an inventory of the foremost appropriate things. Recommendation systems technology may be applied to ease the task of locating the desired things on-line. Method: This study proposes a recommender system allows that permits that allow users to hold out a preference-based cooperative filtering search on rental properties and enables them to refine those preferences based on shallow learning. The room also is visualized using AR and Vuforia. Results: This recommendation approach has been shown to supply additional correct search results. The system was developed as an Internet application victimization Handlebars for front-end and Nodes- ExpressJs for Back-end. Conclusion: The system performs better than existing algorithms and predicts better in memory-based approach. © 2020 The Authors. Published by Elsevier B.V.